The alveolar-capillary network is the largest microvascular bed in the body. Owing to the large alveolar-capillary reserves, extensive loss of the microvascular bed can be tolerated without developing dyspnea at rest;hence pulmonary diabetic microangiopathy is under-recognized clinically. Nevertheless, measurable abnormalities in alveolar-capillary gas exchange develop and can be detected before symptoms appear either in the lung or in other organs. Even modest loss of alveolar microvascular function can be quantified from lung diffusing capacity (DL) and its components: membrane diffusing capacity (Dm) and pulmonary capillary blood volume (Vc), relative to simultaneously measured cardiac output (Qc). From rest to maximal exercise, recruitment of more alveolar capillaries normally causes DL to increase more than 100%. The relationship between DL and Qc correlate strongly with alveolar ultrastructure (alveolar-capillary surface area and blood-gas barrier thickness) and provide accurate indicators of microvascular reserves. DL and Dm measured at a given Qc are significantly reduced in asymptomatic type-1 diabetics with elevated glycosylated hemoglobin (HbAlc), but are almost normal in diabetics of similar disease duration who maintained near-normal HbAlc for 5-6 years;similar impairment is seen in type-2 diabetes and correlates with the prevalence of clinical end-organ complications. Preliminary data lead to these hypotheses: 1) Relationships of DL and Dm to Qc provide quantitative indices for following pulmonary diabetic microangiopathy. 2) Microangiopathy is reversible after a relatively short duration of rigorous glycemic control. We will assess cardiopulmonary function at rest and during graded exercise in patients with recent onset or long-standing type-1 and type-2 diabetes;measurements will be correlated to extrapulmonary indices of microvascular function (nephropathy, retinopathy and neuropathy). Patients will be followed prospectively while glycemic control is optimized and the studies repeated at the end. Results will be analyzed a) longitudinally within and among diabetic groups and with respect to age-matched non-diabetic controls, and b) in cross-sectional comparison to poorly controlled diabetics and normal controls. Our objectives are to establish long-term cohorts to determine the natural progression and potential reversibility of alveolar microvascular dysfunction in diabetes, and to provide prospective reference data for the evaluation of the lung as a route for drug delivery in diabetes. Given the current intense interest in the development of inhaled insulin therapy, there is a growing need to understand the effects of diabetic progression and treatment on lung function.